First, drinking water systems and private wells were identified, mean contaminant levels of DBCP for individual wells were determined from state-derived data, and an evaluation was made of which water system supplied drinking water to each census tract. By using mapping techniques, the authors were able to estimate the geographic percentage of the census tract supplied by each water system. Based on these data, specific weighted averages of arsenic, nitrate, and DBCP by census tract were determined and used in the subsequent ecologic and case-control studies. Mean DBCP levels ranged from 0.0041 to 5.7543 ppb among census tracts. Fourteen (12.8 percent) census tracts had DBCP concentrations in excess of the state's MCL of 1.0 ppb.
There are limitations associated with the studies—for example, no estimate of individual exposure accounts for bottled-water use or other use patterns and whether there is sufficient latency, that is time from first exposure to the development of the disease. But the DBCP studies are serious attempts at defining the historical exposure in greater detail than is generally found in environmental studies, and they should serve as useful models for future investigations.
The findings of the studies are complex. No correlation was found between gastric cancer or leukemia and DBCP exposure in the ecologic analysis. The case-control study did not identify any relationship between gastric cancer and DBCP in drinking water. However, the variable “Hispanic surname” was a risk factor for gastric cancer; Hispanics had a relative risk of gastric cancer of 2.77, compared with non-Hispanics. Hispanics tended to live in areas where the drinking water was more contaminated than did other groups. In addition, farm workers seem to have an increased risk of leukemia, possibly because of occupational exposures —although this will require further study. Dietary factors have not been evaluated. Overall, the case-control study found no association between exposure to DBCP and risk of developing leukemia in persons who live in Fresno County.
The California Department of Health Services (Deane et al., 1989; Swan et al., 1989; Wrensch et al., 1990a,b) has reported on a number of studies designed to assess the basis for an excess of adverse pregnancy outcomes, such as statistically significant spontaneous abortions and birth defects, in Santa Clara County. There were significant concerns in the community that adverse pregnancy outcomes might have occurred as a result of contamination of a single well with trichloroethane (TCA) that had leaked from an underground storage tank owned by a semiconductor manufacturer.